Evaluating the Effectiveness of Prognostic Models in Patients with Cardiogenic Shock Undergoing VA-ECMO Support - Summary - MDSpire

Evaluating the Effectiveness of Prognostic Models in Patients with Cardiogenic Shock Undergoing VA-ECMO Support

  • By

  • Maria Calvo-Barceló

  • Finn Boyhan Irvine

  • Niamh Tierney

  • Shivani Ayyar

  • Taylor Devine

  • Vasileios Panoulas

  • Maria Montegudo-Vela

  • Fernando Riesgo Gil

  • Clara Hernandez Caballero

  • I. Dolores Poveda Pinedo

  • Javier Bautista

  • Jason Van Schoor

  • Donna Hall

  • Sofia Pinto

  • Eftychia Galiatsou

  • Alex Rosenberg

  • April 28, 2026

  • 0 min

Share

Objective:

To evaluate the discriminatory performance of established prognostic scores, including both intensive care and ECMO-specific models, and identify factors associated with survival in patients treated with VA-ECMO for cardiogenic shock.

Key Findings:
  • 265 VA-ECMO runs analyzed; median age 51 years, 34% female.
  • Survival rates: 52% to decannulation, 41% to ICU discharge, 37% to six months, indicating significant mortality.
  • Only SAVE, Alfred, and PREDICT 6 h scores differed significantly between survivors and non-survivors.
  • Discrimination was poor across all models (AUCs 0.52 to 0.68), with the Alfred score showing the best performance (AUC 0.68).
  • Cardiogenic shock aetiology was the only independent predictor of survival.
Interpretation:

Existing prognostic scores for VA-ECMO patients performed poorly, with aetiology being the strongest predictor of survival. General ICU scores were less effective due to their reliance on variables less informative in ECMO patients.

Limitations:
  • Retrospective single-centre analysis may limit generalizability; findings may not apply to other settings.
  • Some prognostic scores could not be calculated due to missing variables, affecting the comprehensiveness of the analysis.
  • Overall mortality exceeded that reported in the ELSO registry, possibly reflecting historical practices and case mix.
Conclusion:

Future prognostic models should incorporate aetiology and other readily available clinical variables to enhance pre-ECMO risk assessment and decision-making, emphasizing the need for improved tools.

Original Source(s)

Evaluating the Effectiveness of Prognostic Models in Patients with Cardiogenic Shock Undergoing VA-ECMO Support

  • by Maria Calvo-Barceló, Finn Boyhan Irvine, Niamh Tierney, Shivani Ayyar, Taylor Devine, Vasileios Panoulas, Maria Montegudo-Vela, Fernando Riesgo Gil, Clara Hernandez Caballero, I. Dolores Poveda Pinedo, Javier Bautista, Jason Van Schoor, Donna Hall, Sofia Pinto, Eftychia Galiatsou, Alex Rosenberg

  • April 28, 2026

Related Content

Masters Athletes Present Unique CV Findings

Joint clinical consensus outlines evaluation and management considerations for arrhythmias, coronary atherosclerosis, aortic dilatation, myocardial fibrosis, and related findings in older competitive athletes.

Earlier Detection of Organ Rejection

CAP review outlines how donor-derived DNA assays can support earlier, less invasive detection of graft injury

  • February 27, 2026

  • 2 min
Top 10 States for Physician-Owned Hospitals

These 10 states make it more practical for physicians to participate in hospital ownership by aligning statutory structure, corporate practice of medicine rules, and population trends.